Abstract Since 1999, Very High spatial Resolution satellite data (Ikonos-2, QuickBird and OrbView-3) represent the surface of the Earth with more detail. However, information extraction by multispectral pixel-based classification proves to have become more complex owing to the internal variability increase in the land-cover units and to the weakness of spectral resolution. Therefore, one possibility is to consider the internal spectral variability of land-cover classes as a valuable source of spatial information that can be used as an additional clue in characterizing and identifying land cover. Moreover, the spatial resolution gap that existed between satellite images and aerial photographs has strongly decreased, and the features used in visual interpretation transposed to digital analysis (texture, morphology and context) can be used as additional information on top of spectral features for the land cover classification. The difficulty of this approach is often to transpose the visual features to digital analysis. To overcome this problem region-based classification could be used. Segmentation, before classification, produces regions that are more homogeneous in themselves than with nearby regions and represent discrete objects or areas in the image. Each region becomes then a unit analysis, which makes it possible to avoid much of the structural clutter and allows to measure and use a number of features on top of spectral features. These features can be the surface, the perimeter, the compactness, the degree and kind of texture. Segmentation is one of the only methods which ensures to measure the morphological features (surface, perimeter...) and the textural features on non-arbitrary neighbourhood. In the pixel-based methods, texture is calculated with mobile windows that smooth the boundaries between discrete land cover regions and create between-class texture. This between-class texture could cause an edge-effect in the classification. In this context, our research focuses on the potential of land cover region-based classification of VHR satellite data through the study of the object extraction capacity of segmentation processes, and through the study of the relevance of region features for classifying the land-cover classes in different kinds of Belgian landscapes; always keeping in mind the parallel with the visual interpretation which remains the reference. Firstly, the results of the assessment of four segmentation algorithms belonging to the two main segmentation categories (contour- and region-based segmentation methods) show that the contour detection methods are sensitive to local variability, which is precisely the problem that we want to overcome. Then, a pre-processing like a filter may be used, at the risk of losing a part of the information. The “region-growing” segmentation that uses the local variability in the segmentation process appears to be the best compromise for the segmentation of different kinds of landscape. Secondly, the features calculated thanks to segmentation seem to be relevant to identify some land-cover classes in urban/sub-urban and rural areas. These relevant features are of the same type as the features selected visually, which shows that the region-based classification gets close to the visual interpretation. The research shows the real usefulness of region-based classification in order to classify the land cover with VHR satellite data. Even in some cases where the features calculated thanks to the segmentation prove to be useless, the region-based classification has other advantages. Working with regions instead of pixels allows to avoid the salt-and-pepper effect and makes the GIS integration easier. The research also highlights some problems that are independent from the region-based classification and are recursive in VHR satellite data, like shadows and the spatial resolution weakness for identifying some land-cover classes. Résumé Depuis 1999, les données satellitaires à très haute résolution spatiale (IKONOS-2, QuickBird and OrbView-3) représentent la surface de la terre avec plus de détail. Cependant, l’extraction d’information par une classification multispectrale par pixel devient plus complexe en raison de l’augmentation de la variabilité spectrale dans les unités d’occupation du sol et du manque de résolution spectrale de ces données. Cependant, une possibilité est de considérer cette variabilité spectrale comme une information spatiale utile pouvant être utilisée comme une information complémentaire dans la caractérisation de l’occupation du sol. De plus, de part la diminution de la différence de résolution spatiale qui existait entre les photographies aériennes et les images satellitaires, les caractéristiques (attributs) utilisées en interprétation visuelle transposées à l’analyse digitale (texture, morphologie and contexte) peuvent être utilisées comme information complémentaire en plus de l’information spectrale pour la classification de l’occupation du sol. La difficulté de cette approche est la transposition des caractéristiques visuelles à l’analyse digitale. Pour résoudre ce problème la classification par région pourrait être utilisée. La segmentation, avant la classification, produit des régions qui sont plus homogène en elles-mêmes qu’avec les régions voisines et qui représentent des objets ou des aires dans l’image. Chaque région devient alors une unité d’analyse qui permet l’élimination de l’effet « poivre et sel » et permet de mesurer et d’utiliser de nombreuses caractéristiques en plus des caractéristiques spectrales. Ces caractéristiques peuvent être la surface, le périmètre, la compacité, la texture. La segmentation est une des seules méthodes qui permet le calcul des caractéristiques morphologiques (surface, périmètre, …) et des caractéristiques texturales sur un voisinage non-arbitraire. Avec les méthodes de classification par pixel, la texture est calculée avec des fenêtres mobiles qui lissent les limites entre les régions d’occupation du sol et créent une texture interclasse. Cette texture interclasse peut alors causer un effet de bord dans le résultat de la classification. Dans ce contexte, la recherche s’est focalisée sur l’étude du potentiel de la classification par région de l’occupation du sol avec des images satellitaires à très haute résolution spatiale. Ce potentiel a été étudié par l’intermédiaire de l’étude des capacités d’extraction d’objet de la segmentation et par l’intermédiaire de l’étude de la pertinence des caractéristiques des régions pour la classification de l’occupation du sol dans différents paysages belges tant urbains que ruraux.

The automated Tree Identification and Delineation Algorithm (TIDA) was developed for application to high spatial resolution digital imagery of Australian native eucalypt forest. The algorithm is based on contiguous, threshold-based spatial clustering of pixels and was designed to cope with the complex asymmetric crowns typical of eucalypts. / To facilitate systematic algorithm evaluation, a forest scene simulation model was created for the simulation of visually realistic remotely sensed images. The model is based on the principles of ray-tracing and the geOll1etric description of scene objects and background. The model simulates the appearance of a forest scene viewed and illuminated from specific directions and under known atmospheric conditions. The distinctive 'modular' structure of eucalypts was represented by modelling crowns as small (branch-scale) spheroids distributed over a larger spheroidal envelope. / Using the simulation model, TIDA performance was evaluated in terms of forest structure (canopy cover, crown cover and canopy structural variability) and the remote sensing environment (view zenith angle, solar zenith angle and aerosol optical thickness). Prior to the evaluation, a methodology was developed for objectively estimating the optimum spatial resolution for TIDA application in a given image. The methodology was based on incremental Gaussian smoothing and exploited TIDA's sensitivity to changes in image spatial resolution. This process demonstrated the importance of individual crown cover, rather than crown size, as the main factor determining the optimum spatial resolution for tree delineation. / Results indicate that TIDA is most suited for application in forests with high canopy cover and high crown cover. The structural complexity of forest canopies, represented by the diameter and overlap of crowns and tree height, had a relatively small impact on TIDA performance. Increasing view zenith angle consistently caused a decrease in TIDA performance. A small phase angle between the sun and sensor produces optimum TIDA performance when both canopy and crown cover is high. As crown or canopy cover decrease, high positive and negative sun zenith angles yield superior TIDA results by decreasing the brightness of the background relative to the canopy and improving the identification of tree peaks. For both dense and sparse canopies, back-scattered radiation from the forest canopy was more suited to automated tree crown delineation than forward-scattered radiation. Imagery acquired under an optically thick atmosphere was found to increase TIDA performance compared to scene illumination under strong direct light. The advantage stemmed from a strengthening of the relationship between geometric and radiometric crown shape. / Through an awareness of limitations imposed by the remote sensing environment, the potential for manipulation of image characteristics, and preferential selection of acquisition conditions, TIDA performance can be optimised to suit various structural forest types. Canopy cover, crown cover, view zenith angle, sun zenith angle, background brightness and image spatial resolution are key criteria in assessing the suitability of automated tree crown delineation as an image interpretation procedure.

With the advent of very high spatial resolution (VHR) satellite, spatial details within the image scene have increased considerably. This led to the development of object-based image analysis (OBIA) for the analysis of VHR satellite images. Image segmentation is the fundamental step for OBIA. However, a large number of techniques exist for RS image segmentation. To identify the best ones for VHR imagery, a comprehensive literature review on image segmentation is performed. Based on that review, it is found that the multiresolution segmentation, as implemented in the commercial software eCognition, is the most widely-used technique and has been successfully applied for wide variety of VHR images. However, the multiresolution segmentation suffers from the parameter estimation problem. Therefore, this study proposes a solution to the problem of the parameter estimation for improving its efficiency in VHR image segmentation. The solution aims to identify the optimal parameters, which correspond to optimal segmentation. The solution to the parameter estimation is drawn from the Equations related to the merging of any two adjacent objects in multiresolution segmentation. The solution utilizes spectral, shape, size, and neighbourhood relationships for a supervised solution. In order to justify the results of the solution, a global segmentation accuracy evaluation technique is also proposed. The solution performs excellently with the VHR images of different sensors, scenes, and land cover classes. In order to justify the applicability of solution to a real life problem, a building detection application based on multiresolution segmentation from the estimated parameters, is carried out. The accuracy of the building detection is found nearly to be eighty percent. Finally, it can be concluded that the proposed solution is fast, easy to implement and effective for the intended applications.

Made available in DSpace on 2014-06-11T19:22:25Z (GMT). No. of bitstreams: 0 Previous issue date: 2006-06-30Bitstream added on 2014-06-13T19:48:44Z : No. of bitstreams: 1 sartori_lr_me_prud.pdf: 1503241 bytes, checksum: 70f9983e4d75d8593ab7f2d397146db7 (MD5) / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) / Imagens multiespectrais de alta resolução espacial podem se constituir em uma fonte de dados adequada para o mapeamento de processos associados ao desenvolvimento de culturas agrícolas, como a detecção de plantas daninhas. A aerofotogrametria convencional e imagens de satélite de alta resolução espacial são alternativas para aquisição dessas imagens. Entretanto, devido ao custo elevado da aquisição destas imagens, tem sido desenvolvido, pelo Departamento de Cartografia da Faculdade de Ciências e Tecnologia da UNESP de Presidente Prudente, um Sistema de Sensoriamento Remoto Aerotransportado (SRA), capaz de oferecer resolução espacial sub-métrica. Este trabalho considerou a hipótese de que a partir de imagens adquiridas pelo Sistema é possível discriminar graus de infestação de plantas daninhas em culturas de café. Para investigar esta hipótese, foi realizado o mapeamento de plantas daninhas utilizando dois diferentes métodos: classificação de imagens multiespectrais (classificação por redes neurais artificiais - RNA) e análise geoestatística (krigagem por indicação com dados indiretos). Os mapas temáticos foram submetidos à análise da qualidade temática. A krigagem por indicação apresentou resultados suavizados e confusos, ao contrário da classificação por RNA, a qual se constituiu num método eficiente para o objetivo proposto, confirmando a hipótese inicial da investigação. / Multispectral images with high spatial resolution can be a suitable data source for the mapping of processes associated to the crop development, as detection of weed. The aerial photogrammetry and satellite image of high spatial resolution are alternatives for the aquisition of these images. However, due to the high cost of these images, a Sistema de Sensoriamento Remoto Aerotransportado - SRA, which is capable of to offer sub-metric spatial resolution has been developed by the Department of Cartography, FCT/Unesp (Presidente Prudente). This work taked into account the hypothesis that is possible to discriminate infestation degree of weed in coffee crop from high spatial resolution multispectral images. To investigate this hypothesis, it was accomplished the mapping using two different methods: multispectral images classification (artificial neural networks classification) and geoestatistics analysis (Indicator kriging with soft data). The thematics maps were submitted to the analysis of thematic quality. The indicator kriging showed smoothed and confused results instead of the artificial neural networks classification, whose results were efficient for the purpose, confirming the initial hypothesis of the investigation.

<p>   </p> <p>Ambient Mass Spectrometry Imaging (MSI) is a powerful analytical tool in biomedical research that enables simultaneous label-free spatial mapping of hundreds of molecules in biological samples under native conditions. Nanospray desorption electrospray ionization (nano-DESI) is an emergent ambient MSI technique developed in 2010 that uses localized liquid extraction of molecules directly from surfaces. Like other liquid-extraction based techniques, nano-DESI relies on gentle removal of molecules from surfaces and soft ionization. High sensitivity and spatial resolution, versatility of the solvent composition, which may be used to tailor the extraction and ionization of selected molecules, quantification capabilities at the single-pixel level as well as compensation for matrix effects by adding a known standard to the solvent, and online derivatization are key features of nano-DESI MSI that position it as a unique analytical tool for studying biological systems. </p> <p>Despite the advantages that nano-DESI provides, there are still challenges associated with the structural characterization, extraction, and detection of certain molecular classes. Therefore, my dissertation research has focused on addressing these analytical challenges by developing innovative approaches that substantially enhance the performance of the nano-DESI technique in the study of complex biological systems. </p> <p>In this thesis, a systematic study of the solvent composition is carried out to aid in the detection of neutral lipids such as triglycerides thereby expanding the molecular coverage of nano-DESI experiments. Taking advantage of the versatility of the solvent composition, I developed an approach for the online derivatization of unsaturated lipids into lipid hydroperoxides using the reaction of singlet oxygen with C=C bonds. This method further expands the specificity of nano-DESI MSI by enabling the detection and imaging of positional lipid isomers. To aid in the analysis of complex mixtures and provide additional structural information in the form of collision cross sections, coupling of nano-DESI with a drift-tube ion mobility spectrometry is also reported along with examples of the powerful capabilities of this platform. Lastly, nano-DESI MSI is used to address the complexity in the analysis of individual skeletal muscle fibers. This collaborative project involves the development of a robust image registration approach of immunofluorescence imaging and high-spatial resolution nano-DESI MSI to obtain accurate chemical maps specific to each fiber type. The developments described in this thesis are key to understanding the dynamic metabolic processes on a molecular level with an unprecedented specificity and sensitivity.</p> <p>  </p>

Lignocellulosic plants are the most abundant source of terrestrial biomass and are one of the potential sources of renewable energy that can replace the use of fossil fuels. For a country such as Sweden, where the forest industry accounts for 10% of the total export, there would be large economical benefits associated with increased biomass yield. The availability of research on wood development conducted in conifer tree species, which represent the majority of the forestry in Sweden, is limited and the majority of research has been conducted in model angiosperm species such as Arabidopsis thaliana. However, the large evolutionary distance between angiosperms and gymnosperms limits the possibility to identify orthologous genes and regulatory pathways by comparing sequence similarity alone. At such large evolutionary distances, the identification of gene similarity is, in most cases, not sufficient and additional information is required for functional annotation. In this thesis, two high-spatial resolution datasets profiling wood development were processed; one from the angiosperm tree Populus tremula and the other from the conifer species Picea abies. These datasets were each published together with a web resource including tools for the exploration of gene expression, co-expression and functional enrichment of gene sets. One developed resource allows interactive, comparative co-expression analysis between species to identify conserved and diverged co-expression modules. These tools make it possible to identifying conserved regulatory modules that can focus downstream research and provide biologists with a resource to identify regulatory genes for targeted trait improvement. / Lignocellulosa är den vanligast förekommande källan till markburen biomassa och är en av de förnybara energikällor som potentiellt kan ersätta användningen av fossila bränslen. För ett land som Sverige, där skogsindustrin som står för 10 \% av den totala exporten, skulle därför en ökad produktion av biomassa kunna ge stora ekonomiska fördelar. Forskningen på barrträd, som utgör majoriteten av svensk skog är begränsad och den huvudsakliga forskningen som har bedrivits på växter, har skett i modell organismer tillhörande gruppen gömfröiga växter som till exempel i Arabidopsis thaliana. Det evolutionära avståndet mellan gömfröiga (blommor och träd) och nakenfröiga (gran och tall) begränsar dock möjligheten att identifiera regulatoriska system mellan dessa grupper. Vid sådana stora evolutionära avstånd krävs det mer än att bara identifiera en gen i en modellorganism utan ytterligare information krävs som till exempel genuttrycksdata. I denna avhandling har två högupplösta experiment som profilerar vedens utveckling undersökts; ett från gömfröiga träd Populus tremula och det andra från nakenföriga träd (barrträd) Picea abies. Datat som behandlats har publicerats tillsammans med webbsidor med flera olika verktyg för att bland annat visa genuttryck, se korrelationer av genuttryck och test för anrikning av funktionella gener i en grupp. En resurs som utvecklats tillåter interaktiva jämförelser av korrelationer mellan arter för att kunna identifiera moduler (grupper av gener) som bevaras eller skilts åt mellan arter över tid. Identifieringen av sådana bevarade moduler kan hjälpa att fokusera framtida forskning samt ge biologer en möjlighet att identifiera regulatoriska gener för en riktad förbättring av egenskaper hos träd.

Analysis of forest canopy density based on textural features of high resolution imagery and airborne laser scanning data Abstract The objective of this thesis is to assess the forest canopy density in the Šumava Mountains, Czech Republic. The spruce forests in this area have been suffering from the bark beetle outbreak for almost 20 years resulting in a mixture of dead and young trees, mature forest stands and peat bogs. The canopy density was evaluated using a very high spatial resolution panchromatic imagery and low point density LiDAR, combined with an object oriented approach. The classification based on three GLCM texture measures (contrast, entropy and correlation), which were derived from the image objects, resulted in a kappa index of accuracy of 0.45. Adding the information from the LiDAR data, the accuracy of the classification improved up to 0.95.

Les activités anthropiques dans le bassin méditerranéen sont en forte évolution. Dans les zones agricoles, cette croissance entraîne des évolutions considérables de l'occupation du sol. Cette activité agricole exerce un impact majeur sur le fonctionnement hydrologique des paysages qui n'est identifiable qu'à une échelle bien plus large, sur plusieurs dizaines de km². Cette thèse se concentre sur la classification de l'occupation du sol sur une large étendue à partir d'une image monodate à haute résolution spatiale (SPOT6/7).Dans ce contexte, les données d'apprentissage sont collectées par des enquêtes terrain, par conséquent, elles sont très limitées. Les méthodes d'apprentissage supervisées sont généralement utilisées, en supposant que la distribution des classes est stable sur toute l'image. Cependant, en pratique, on constate une distorsion des distributions des classes (apparition de nouvelles classes, disparition de classes). Ce problème, intitulé "datashift", se produit souvent sur des larges étendues. Ainsi le modèle construit sur les données d'apprentissage initiales s'avère sous optimal pour la classification de l'image entière. Pour atténuer ce problème, les techniques d'apprentissage actif définissent un ensemble d'apprentissage efficace, en l'adaptant itérativement par l'ajout des données non labellisées les plus informatives. Ces techniques permettent d'améliorer le modèle de classification tout en conservant un petit ensemble d'apprentissage initial. L'échantillonnage se base généralement sur deux métriques : l'incertitude et la diversité.Dans cette thèse, nous montrons l'apport des techniques d'apprentissage actif pour la cartographie de l'occupation du sol en milieu agricole, en proposant un échantillonnage adapté par parcelle.L'apport des méthodes d'apprentissage actif est validé par rapport à une sélection aléatoire des parcelles. Une métrique de diversité basée sur l'algorithme Meanshift a été proposée.Dans un deuxième temps, nous avons traité le sous-problème du "datashift" qui est l'apparition de nouvelles classes. Nous avons proposé de nouvelles métriques de diversité basées sur l'algorithme Meanshift et les Fuzzy k-means ainsi qu'une nouvelle stratégie de sélection des données adaptées à la détection de nouvelles classes.Dans la dernière partie, nous nous sommes intéressés aux contraintes spatiales induites par les observations sur terrain et nous avons proposé une stratégie de labellisation par points de vue qui permet de diminuer largement les coûts humains d'observations terrain tout en gardant de bonnes précisions de classification ainsi que la découverte des nouvelles classes.Les méthodes proposées ont été testées et validées avec une image multispectrale SPOT6 à 6m de résolution sur le bassin versant de Lebna, Cap-Bon, Tunisie. / Anthropogenic activities in the Mediterranean are in strong evolution. In agricultural areas, this growth leads to considerable changes in land cover. This agricultural activity has a major impact on the hydrological functioning of the landscapes which can be only identified on a wide scale, over several tens of km². This thesis focuses on the land cover classification on wide area from a high spatial resolution monodate image (SPOT6/7).In this context, the learning data are collected by field surveys, therefore they are very limited. Supervised learning methods are generally used, assuming that the class distribution is stable over all the image. However, in practice, there is a class distributions distortion (new classes appear, classes disappear). This problem, called "datashift", always occurs over wide areas. Thus, the model constructed on the initial learning data is sub-optimal for the classification of the entire image. To lessen this problem, active learning techniques define an effective learning set, by iteratively adapting it by adding the most informative unlabeled data. These techniques improve the classification model while retaining a small initial learning set. Sampling is generally based on two metrics: uncertainty and diversity.In this thesis, we show the contribution of active learning techniques for the land cover mapping in agricultural environment, proposing a suitable sampling per parcel.The active learning methods contribution is validated respectively to a random selection of parcels. A diversity metric based on the Meanshift algorithm has been proposed.Secondly, we treated the sub-problem of the "datashift" which is the appearance of new classes. We proposed new metrics of diversity based on the Meanshift algorithm and Fuzzy k-means as well as a new data selection strategy adapted to the detection of new classes.Finally we were interested in the spatial constraints induced by the field observations and we proposed a strategy of labeling by stand points which make it possible to greatly reduce the human costs for field observations while maintaining good classification precisions as well as the discovery of new classes.The proposed methodologies were tested and validated on a multispectral SPOT6 image with 6m resolution on the Lebna watershed, Cap-Bon, Tunisia.

Nos centros urbanos, as precipitações sempre estiveram ligadas a problemas como inundações e propagação de doenças. Para solucioná-los, é comum a realização de obras hidráulicas nos sistemas de drenagem urbanos. Para tanto, deve-se conhecer as condições da bacia hidrográfica e as consequências que qualquer alteração no ambiente pode causar. Portanto, modelos hidrológicos são utilizados na previsão do comportamento das bacias frente a eventos de precipitação, aumentando a eficácia das obras e diminuindo os riscos associados a estas. Para o uso de modelos, são necessários diversos parâmetros hidrológicos referentes à bacia, tais como área de drenagem, comprimento e declividade dos talvegues, tipo de cobertura de solo etc. Com o avanço da tecnologia, a determinação destes torna-se cada vez mais precisa, bem como os modelos utilizados, trazendo o Sistema de Informações Geográficas (SIG) e o sensoriamento remoto como poderosas ferramentas de apoio a estudos hidrológicos. Neste trabalho, aplicou-se o processo de classificação automática supervisionada pelo método da Análise Orientada a Objeto a uma imagem de satélite de alta resolução da bacia hidrográfica do córrego do Gregório, para caracterizar sua cobertura de solo e determinar os parâmetros hidrológicos número de deflúvio (CN, pelo método do SCS), grau de vegetação (PP), área (A), comprimento (L) e declividades dos talvegues (S) das sub-bacias que compõem a bacia, para as quais os resultados obtidos foram bastante satisfatórios. Por fim, atualizou-se o modelo hidrológico EESC (1993), referente ao sistema de macrodrenagem de São Carlos, obtendo-se hidrogramas finais com diferenças, em relação ao modelo original, de até 33,96% para vazão de pico (Qp), 77,78% para tempo de pico (tp) e 29,86% para volume total de escoamento. / In urban centers, precipitation always been related to problems such as floods and spread of disease. To solve them, it is common to make hydraulic interventions in the urban drainage systems. For this, it is necessary to know the conditions of the watershed and the consequences that any change in the environment can cause. Therefore, hydrological models are used to predict the river behavior in opposite to precipitation events, increasing the efficiency of the hydraulic interventions and reducing the associated risks to these. For the use of models, it is necessary to have several hydrological parameters related to the basin such as drainage area, river length, slope of the thalweg, type of soil cover etc. Trough the technological advancement, the parameter determination becomes more accurate as well as the models, and the Geographic Information System (GIS) and remote sensing appear as powerful tools to support hydrological studies. In this study, we have applied the automatic supervised classification process by the Object-Oriented Analysis method to a high resolution satellite image of the córrego do Gregório watershed, to classify soil coverage and to determine the hydrological parameters curve-number (CN by the SCS method), vegetation degree (PP), area (A), length (L), and slope of thalwegs (S) of the sub-basins of the córrego do Gregório watershed, for which the results were quite satisfactory. Finally, a hydrological model for the São Carlos macrodrainage system called EESC model (1993) was updated with the new parameters, obtaining final hydrographs with differences from the original model up to 33.96% for peak discharge (Qp), 77.78% for peak time (tp) and 29.86% for total volume of runoff.

Depuis une quinzaine d’années, l’émergence des données de télédétection à Très Haute Résolution Spatiale (THRS) et la démocratisation des Systèmes d’Information Géographique (SIG) aident à répondre aux nouveaux besoins croissants d’informations spatialisées. Le développement de nouvelles méthodes de cartographie offre une opportunité pour comprendre et anticiper les mutations des surfaces terrestres aux grandes échelles, jusqu’ici mal connues. En France, l’emploi de bases de données spatialisées sur l’occupation du sol à grande échelle (BD Ocsol GE) est devenu incontournable dans les opérations courantes de planification et de suivi des territoires. Pourtant, l’acquisition de ce type de bases de données spatialisées est encore un besoin difficile à satisfaire car les demandes portent sur des productions cartographiques sur-mesure, adaptées aux problématiques locales des territoires. Face à cette demande croissante, les prestataires réguliers de ce type de données cherchent à optimiser les procédés de fabrication avec des techniques récentes de traitements d’image. Cependant, la Photo-Interprétation Assistée par Ordinateur (PIAO) reste la méthode privilégiée des prestataires. En raison de sa grande souplesse, elle répond toujours au besoin de cartographie aux grandes échelles, malgré son coût important. La substitution de la PIAO par des méthodes de production entièrement automatisées est rarement envisagée. Toutefois, les développements récents en matière de segmentation d’images peuvent contribuer à l’optimisation de la pratique de la photo-interprétation. Cette thèse présente ainsi une série d’outils (ou modules) qui participent à l’élaboration d’une assistance à la digitalisation pour l’exercice de photo-interprétation d’une BD Ocsol GE. L’assistance se traduit par la réalisation d’un prédécoupage du paysage à partir d’une segmentation menée sur une image THRS. L’originalité des outils présentés est leur intégration dans un contexte de production fortement contraint. La construction des modules est conduite à travers trois prestations cartographiques à grande échelle commandités par des entités publiques. L’apport de ces outils d’automatisation est analysé à travers une analyse comparative entre deux procédures de cartographie : l’une basée sur une démarche de photo-interprétation entièrement manuelle et la seconde basée sur une photo-interprétation assistée en amont par une segmentation numérique. Les gains de productivité apportés par la segmentation sont, évalués à l’aide d’indices quantitatifs et qualitatifs, sur des configurations paysagères différentes. À des degrés divers, il apparaît que quelque soit le type de paysage cartographié, les gains liés à la cartographie assistée sont substantiels. Ces gains sont discutés, à la fois, d’un point de vue technique et d’un point de vue thématique dans une perspective commerciale. / Over the last fifteen years, the emergence of remote sensing data at Very High Spatial Resolution (VHRS) and the democratization of Geographic Information Systems (GIS) have helped to meet the new and growing needs for spatial information. The development of new mapping methods offers an opportunity to understand and anticipate land cover change at large scales, still poorly known. In France, spatial databases about land cover and land use at large scale have become an essential part of current planning and monitoring of territories. However, the acquisition of this type of database is still a difficult need to satisfy because the demands concern tailor-made cartographic productions, adapted to the local problems of the territories. Faced with this growing demand, regular service providers of this type of data seek to optimize manufacturing processes with recent image-processing techniques. However, photo interpretation remains the favoured method of providers. Due to its great flexibility, it still meets the need for mapping at large scale, despite its high cost. Using fully automated production methods to substitute for photo interpretation is rarely considered. Nevertheless, recent developments in image segmentation can contribute to the optimization of photo-interpretation practice. This thesis presents a series of tools that participate in the development of digitalization assistance for the photo-interpretation exercise. The assistance results in the realization of a pre-cutting of the landscape from a segmentation carried out on a VHRS image. Tools development is carried out through three large-scale cartographic services, each with different production instructions, and commissioned by public entities. The contribution of these automation tools is analysed through a comparative analysis between two mapping procedures: manual photo interpretation versus digitally assisted segmentation. The productivity gains brought by segmentation are evaluated using quantitative and qualitative indices on different landscape configurations. To varying degrees, it appears that whatever type of landscape is mapped, the gains associated with assisted mapping are substantial. These gains are discussed both technically and thematically from a commercial perspective.